[VoL. 8 
14 ANNALS OF THE MISSOURI BOTANICAL GARDEN 
one notes the disappearance of some of the smaller spots of 
the previous year, and often the number of new infections is 
considerable. 
If diseased cotton stalks are left standing in the field, few 
or no evidences of the fungus are apparent on the roots the 
following March. However, some of the more interested 
growers claim to have observed mats of the fungus turned 
over by the plow when bedding the land. I have been unable 
to obtain such material for study. Ав already indicated, the 
reappearance of the larger spots, particularly, is a strong 
indication of the persistence of the mycelium in the soil. This 
leaves out of consideration the influence of the conidial stage, 
discussed below, in the persistence of the organism in the 
same area during successive years. 
On lifting wilted stalks of cotton, or stalks recently dead, it 
is found, from the most favorable material, that the roots are 
closely invested with a einnamon-buff! felt of hyphae in which 
strands are conspicuous. The fungus may involve the small- 
est rootlets, and in addition, the strands of hyphae penetrate 
the soil and apparently extend considerable distances. The 
larger soil strands are somewhat darker, often cinnamon- 
colored. In a badly infected area the strands of hyphae may 
be found in any lump of soil. Pammel describes the mycelium 
as brown in color, and Shear as ‘‘dirty yellow, whitish when 
young.’’ In the early stages of development on the host, I 
find the mycelium pale buff, becoming cinnamon-brown as 
strands are formed. 
In September, 1915, the conditions were particularly favor- 
able at Paris, Texas, for late-season infections, so that by ex- 
amining the roots of many plants taken at the periphery of 
a diseased area, but themselves apparently healthy, compara- 
tively early stages of infection were observed. In all cases a 
depression of the bark pointed out the area of penetration of 
the fungus on the main root. The observations also demon- 
strated clearly that the attack may be either what I shall 
designate centripetal or centrifugal. In the former case the 
infection converges upon the main roots from a few or many 
small laterals, while in the latter the main root may be com- 
*Ridgway’s ‘Color standards and nomenclature’ has been employed in the 
determination of all colors referred to in this paper. 
